Giovanni Braccini

Detection of biliary complications after orthotopic liver transplantation with MR cholangiography

To assess the diagnostic value of magnetic resonance cholangiography (MRC) when evaluating biliary complications in the follow-up of liver transplant patients. One hundred and thirteen patients prospectively underwent MR imaging and MR cholangiography at 1.5-T unit after orthotopic liver transplantation (OLT). After the acquisition of axial T1- and T2-weighted sequences, MRC involved a coronal, non breath-hold, respiratory-triggered, fat-suppressed, two-dimensional, thin-slab, heavily T2-weighted fast spin-echo sequence, and coronal breath-hold, thick-slab, single-shot T2-weighted sequences. The images and maximum intensity projections were evaluated by two readers in order to determine biliary anatomy and the presence of complications, whose final diagnosis was based on endoscopic retrograde cholangiography (ERC) in 50 patients, percutaneous trans-hepatic cholangiography (PTC) in five, and by integrating clinical follow-up with ultrasound and MR findings in 58 cases. MRC had a sensitivity of 93%, a specificity of 92%, a positive predictive value of 86%, a negative predictive value of 96%, and a global diagnostic accuracy of 93% in detecting all types of biliary complications in OLT patients. MRC is a reliable technique for detecting post-OLT biliary complications. We now restrict the use of ERC to patients for whom therapeutic procedures are advocated or whose MRC results are equivocal.

MR virtual endoscopy of the pancreaticobiliary tract

PURPOSE To evaluate the feasibility of surface-rendered magnetic resonance virtual endoscopy (MRVE) of magnetic resonance cholangiopancreatography (MRCP) data sets. We retrospectively reviewed MR cholangiopancreatography data sets of 120 patients with biliary stone (n=40), inflammatory ampullary stenosis (n=12), pancreatic tumor (n=8), cholangiocarcinoma (n=7), stenosis of surgical bilio-enteric anastomosis (n=4), extrinsic localized common bile duct stenosis (n=2), ampullary carcinoma (n=2), pancreatic duct stone (n=1), tumor of the gallbladder (n=1), and normal pancreaticobiliary tree (n=43). MRVE views were generated with Navigator software. Segmentation of the acquired data sets was performed with a thresholding technique. Navigation sequences were simulated through the entire biliary tract. MRVE was obtained in 27 (63%) of the 43 normal patients. Endoscopic views were generated in all 77 patients with partial or complete obstruction of the pancreaticobiliary tree. Among these, three groups of patterns were identified: 36 (47%) endoluminal masses (polyp-like masses), 17 (22%) luminal stenoses, 24 (31%) luminal occlusion. In 29 cases, hole artifacts through the internal wall were observed and interpreted as mistakes of segmentation. MRVE proved to show the internal anatomy of the biliary tract and endoluminal changes due to pathological condition. Further investigations are needed to test the usefulness and the potentialities of this technique.

Incidentally discovered adrenal masses: evaluation with Gadolinium enhancement and fat-suppressed MR imaging at 0.5 T

The purpose of the study is to evaluate the ability of Gd-enhancement and fat-suppressed MR imaging operating at midfield strength to characterize incidentally discovered adrenal masses. Sixty patients with 72 adrenal masses incidentally discovered during US or CT exams were studied with a 0.51 MR unit following clinical and laboratory evaluation. After Gd-DTPA intravenous administration a modified three-point Dixon technique was performed in all patients. This technique provided three images sets: conventional T1-weighted SE images, fat-suppressed T1-weighted images and water-suppressed T1-weighted images. Diagnosis was established by means of surgery (11 lesions), fine-needle biopsy (21 lesions) and stability on ultrasonographic follow-up for at least 1 year (range, 12-87 months) from adrenal lesion discovery (40 masses). In most of adenomas (n = 55) an homogeneous enhancement was observed on postcontrast T1WI; however, 15 out of these lesions showed a small focal spot of high intensity in Gd-enhanced fat-suppressed images. On the contrary, malignant conditions (n = 6) and pheochromocytoma (n = 1), all had inhomogeneous signal intensities which were relatively higher after Gadolinium injection as compared with the liver. The fat suppression technique demonstrated areas of bright signal intensity related to high vascularity. The performance of three observers in order to differentiate malignant from benign conditions showed sensitivity, specificity, diagnostic accuracy, positive and negative predictive values of 100, 88.5, 90, 50 and 100% on the basis of gadolinium enhancement only, by utilizing the Dixon technique. In conclusion, although Gd-enhancement and fat-suppressed sequence helped correctly differentiate among the groups of incidentally discovered adrenal masses, the degree of overlap suggests that it is still difficult to characterize individual patients. However, the modified three-point Dixon technique after contrast material administration appears to be a further capability of midfield MRI in the characterization of adrenal tissue.

Diagnosis of adrenal adenoma: value of central spot of high-intensity hyperintense rim sign and homogeneous isointensity to liver on gadolinium-enhanced fat-suppressed spin-echo MR images.

Eighty‐nine patients with 108 adrenal masses, either adenomas (n = 88) or malignant lesions (n = 20), underwent magnetic resonance imaging (MRI) of the abdomen at 0.5 T for the purpose of determining whether adrenal adenomas could be differentiated from malignant lesions on gadolinium‐enhanced fat‐suppressed T1‐weighted spin‐echo (SE) images (Gd‐E FS T1WI) and on T2‐weighted SE images. The imaging protocol included conventional unenhanced SE T1‐ and T2‐weighted sequences and Gd‐E FS T1WI. Three observers independently evaluated signal intensity on unenhanced and enhanced images and also the presence of structures of high signal intensity in the outer margin [hyperintense rim sign (HRS)] or in the center [hyperintense central spot (HCS)] of the adrenal masses. Forty‐one (46.5%) of 88 adenomas were homogeneously isointense to liver in unenhanced and enhanced T1‐weighted sequences and in T2WI. HCS and HRS were observed in 33/88 (37.5%) and 15/88 (17%) adenomas, respectively, on Gd‐E FS T1WI; in contrast, these signs were never revealed in any case of malignant lesions. Sensitivity, specificity, positive predictive value, negative predictive value, and overall accuracy in classifying lesions as suggestive of adenoma were 93%, 90%, 98%, 75%, and 93%, respectively. Visual evaluation of details of tumor structures on Gd‐E FS T1WI allows good characterization of adrenal masses. HCS, HRS, and homogeneous isointensity to liver are characteristic signs of adrenal adenomas. J. Magn. Reson. Imaging 1999; 9:304–310.

MR cholangiopancreatography: value of axial and coronal fast Spin-Echo fat-suppressed T2-weighted sequences

OBJECTIVE To compare axial and coronal fast Spin-Echo fat-suppressed T2-weighted sequences with three-dimensional (3D) maximum intensity projection (MIP) images in patients with suspected pancreaticobiliary obstruction. MATERIAL AND METHODS MR cholangiopancreatography (MRCP) was performed in 108 consecutive patients with a non-breath-hold, fat-suppressed, 2D, heavily T2-weighted fast spin-echo sequence in coronal plane. Axial T1- and T2-weighted images were previously obtained. In addition, 3D reconstructions of the coronal images were analysed separately by using a MIP algorithm. Both two-dimensional (2D) (axial and coronal) and 3D MIP images were separately evaluated by two readers in conference and their results were compared with that of endoscopic retrograde cholangiopancreatography, percutaneous trans-hepatic cholangiography, surgery and/or imaging follow-up. Statistical analysis of 2D and 3D MRCP images in diagnosing the level and probable cause of pancreaticobiliary obstruction were separately calculated. RESULTS 106/108 of MRCP examinations were judged diagnostic by the two reviewers for adequacy of visualisation of the biliary and pancreatic ducts. Sensitivity, specificity, positive predictive value, negative predictive value and global diagnostic accuracy of 2D (axial and coronal) and 3D MRCP images in diagnosing the pancreaticobiliary obstruction were 94 and 57%,, 95 and 93%, 97 and 92%, 91 and 60%, 94 and 72% respectively. CONCLUSION Our results do indicate a higher global accuracy for axial and coronal fast Spin-Echo fat-suppressed T2-weighted sequences versus 3D MIP images in diagnosis of the level and probable cause of pancreaticobiliary obstruction and stress the limitations of 3D images in depiction of small intraductal pathology such as calculi and biliary neoplastic.

An intelligent system for the diagnosis of complex images.

An intelligent system suitable to perform a computer aided diagnosis of complex images should have a knowledge base containing all information related both to the images to be interpreted and to their symbolic description. In this paper, a system able to classify unknown medical digital images into four classes is proposed (searched pathology recognized, searched pathology absent, different pathology from the searched one recognized, unknown pathology). A main component of this system is a knowledge base that, startling from information deduced from sample images, can be processed to create synthetic reference models that, in turn, permit the interpretation of real scenes. The system has been tested on digitized plain film of the thorax, in order to perform a computer-aided diagnosis of pneumothorax cases.

Usefulness and safety of pirenzepine in double-contrast study of upper gastrointestinal tract: comparison with scopolamine methylbromide.

To evaluate usefulness of pirenzepine, a selective M1 antimuscarinic drug, for diagnostic doublecontrast study of the upper gastrointestinal tract, pirenzepine and scopolamine methylbromide (SMB) were compared in a single blind randomized trial. Seventy consecutive patients were enrolled in the study. Artifacts, bowel distention, painting of stomach and duodenal bulb, and global quality of the images were blindly evaluated by four independent observers by means of a numerical score (1–4). Under SMB slightly but significantly better results for stomach were scored (3.1±.7 vs. 2.7±7, p< 0.01). No differences were found in the study of the duodenal bulb. Heart rate and rhythm during the study were recorded by electrocardiogram (ECG). SMB induced tachycardia in all patients (from 77±20 to 117±28 beats/min, p < 0.01) while pirenzepine did not (from 77±16 to 81±23, p = NS). After SMB, two patients exhibited faintness, and some patients complained of visual accommodation defects, dryness of the mouth, and dizziness. Thus, pirenzepine provides good results in double-contrast studies (equal to SMB), while presenting no adverse effects. It could be proposed as a first choice hypotonic agent in upper gastrointestinal examination.

MR virtual endoscopy of the pancreaticobiliary tract: a feasible technique?

To evaluate the feasibility of magnetic resonance (MR) virtual endoscopy of the pancreaticobiliary tract by using MR cholangiopancreatography (MRCP) data sets as source images, we retrospectively reviewed MRCP data sets of 120 patients with Navigator software (GE/Medical Systems, Milwaukee, WI) that allowed display of inner views by surface rendering the internal wall of the bile ducts with simulated light and shadow.

Sonographic Volume Processing of Malignant Prostatic Lesions

The aim of this study was to investigate the possibility of a precise definition of the volumetric and spatial features of malignant prostatic lesions needing radiation therapy and a later follow-up. The modeling procedure is mainly composed of three sequential phases: a) automatic acquisition of time sequences of 2D echotomograms; b) 3D reconstruction of images; c) generation of a spatial geometrical model.

Real-time volume rendering of MRCP: clinical applications

MR-cholangiopancreatography (Signa Contour 0.5T; GE/Medical Systems, Milwaukee, WI) data sets of 156 patients, obtained with a 2D T2-weighted FSE sequence, in the coronal plane, were volume rendered (Advantage Windows 3.1; GEMS) independently by two radiologists, that were asked to define the range of signal intensitics in which the signal of the pancreaticobiliary system was included and to rank the quality of native images and volume renderings. Patients had biliary stones (n=47), inflammatory ampullary stenoses (n=18), pancreaic tumors (n=12), surgical bilio-enteric anastomoses (n=19), ampullary carcinomas (n=2), pancreatic duct stone (n=1), cholangiocarcinoma (n=3) and normal pancreaticobiliary tree (n=54). Good quality volume renderings of the bile ducts were obtained for at least a maximum diameter of 1.5 mm. The quality rank agreement between volume rendering and native images was excellent (k=0.94). The correlation between the observers for the setting the signal intensity range was excellent and statistically significant (p<0.001). The correlation between the observers for the time of volume rendering was not statistically significant. Bilary stones could be displayed in 32/47 (68%) cases. The pancreatic duct stones was displayed as well. Inflammatory ampullary stenoses were detected in all cases (100%). In case of pancreatic tumors, cholangiocarcinomas and ampullary carcinomas volume rendering allowed to identify the site of stenosis. In surgical bilio-enteric anastomoses volume rendering was helpful to display the residual bilary tract, the site of anastomosis and the enteric tract. Volume rendering could be a reliable and efficient tool for the study of the anatomy and pathological changes of the pancreaticobiliary tract.